Post on 21-Dec-2015
PROSPEC: Using an information technology-supported patient-centered
intervention to reduce disparities in prostate cancer screening discussions
John H. Holmes, PhD1, Ransom Weaver2,Carmen Guerra, MD, MSCE3, Dominick Frosch, PhD4,
Ruthann Auten, AB1, Maryte Curran1, Judy Shea, PhD3, and Robert Hornik, PhD5
1Center for Clinical Epidemiology and Biostatistics, University of Pennsylvania School of Medicine2Ransom Weaver Visual Media and Interactive Design
3Division of General Internal Medicine, University of Pennsylvania School of Medicine 4Division of General Internal Medicine & Health Services Research, UCLA School of Medicine
5Annenberg School of Communication, University of Pennsylvania
CCEBCCEB
Background
• Prostate cancer screening among asymptomatic men is controversial – Treatment can reduce quality of life – Unclear that treatment reduces prostate cancer
mortality.
• Current guidelines – Practitioners should discuss screening with patients– Decision to be screened should be one shared
between patient and physician– This discussion and shared decision making is not as
prevalent when the patient is African American, and thus at higher risk of prostate cancer.
Objective
• To develop and evaluate a patient-centered intervention designed to address the racial disparity in the incidence and quality of prostate cancer screening-related discussions.
• The purpose of the intervention is to provide men with knowledge and skills needed to initiate and participate in a discussion about prostate cancer screening with their physician.
Methods
• The intervention design was informed by preliminary qualitative research and an evidence-based ontology implemented in Protégé, both guided by the Systems Model of Clinical Preventive Care.
• Intervention prototypes developed and evaluated iteratively by lay focus groups.
Conceptual framework Systems Model of Clinical Preventive Care§
§ Walsh JME and McPhee SJ. Health Ed Q 19(2):157-175 (1992)
Three Sources of Information
• Evidence-based ontology
• Community-based interviews
• Physician interviews
Evidence-based ontology
• ~800 items– Academic literature– Popular press– Transcripts
• Items coded in Protégé§ using the Systems Model as a framework
• Relationships between classes of items elucidated
§ protege.stanford.edu
High-level Ontology
Focusing on predisposing factors…
Community InterviewsRecruitment and Methodology
• 18 African American males and 14 Caucasian males recruited using the UPHS and a small local paper advertisement
• Eligibility: Age 40-75, No prior prostate history
• Semi-structured interview – Questions about health care, thoughts and opinions
about prostate cancer and prostate cancer screening, health discussions, and computer usage
Physician InterviewsRecruitment and Methodology
• Participants
- 17 completed interviews9 Internal Medicine faculty3 Family Medicine faculty5 CCA physicians
- 2 pending interviews
- 10 interviews currently being analyzed• Methods
- Semi-structured in-depth interview
- Chart-stimulated recall
Implications for the intervention• Model desirable shared decision making behavior in a
realistically presented setting
• Tailor content and presentation level to patient’s educational level
• Address the patient’s specific concerns about PC and PC screening
• Provide a hard copy artifact for the patient to take into the exam room and use as discussion aid
• Provide a hard copy artifact for the patient to take home
Intervention design
• The intervention uses elements from Cegala’s PACE model of physician-patient communication§
• Intervention scripts were developed by the entire research team
• The intervention incorporates tailoring on physician race and gender and patient race
• Professionally-acted, simulated clinical encounter that models good physician-patient communication
• The scenario is punctuated by narration and interaction to reinforce key points
§ Cegala et al: Patient Educ Counseling 41: 209-222 (2000)
Modified PACE Model
• Ask questions
• Check understanding
• Express concerns
Tailoring on physician
Tailoring on patient
Simulated encounter
Reinforcement
Next steps
• Evaluation of intervention design by focus groups and individual interviews– Usability– Usefulness– What’s missing (or should be!)
• Evaluation of intervention efficacy by quasi-experimental pre-post intervention study in primary care
Conclusion
• Participatory design is key- don’t design patient behavioral interventions in a vacuum!
• The intervention is the first to use informatics and information technology in focusing on racial disparities in discussing prostate cancer screening.
This study was sponsored by the Center for Population Health and Health Disparities
at the University of Pennsylvania under
Public Health Services Grant P50-CA105641